mapped on a tranverse mercator projection. The data may be displayed at the largest 

 scale of 1:7,500,000 along a swath of 1560 n.mi. astride the trace of the satellite sub- 

 point. The infrared data from the HRIR sensor contains internal calibration values using 

 deep space as the level of power and a temperature-controlled sensor housing monitored 

 by thermistors calibrated to within 0.2°C as a calibration source. The noise equivalent 

 temperature difference is less than 1°C r.m.s. over the full temperature range. The 

 original analog signal covers the range of effective radiation temperatures from 210°-310°K, 

 Within this range the global sea surface emits radiation in only the 273°K (0°C) to 303°K 

 (30°C) portion and in any particular geographic region in only a fraction of that range. 

 The HRIR data can be displayed as imagery with 16 gray shade levels, each representing a 

 temperature interval of 1 .6°C. in this way the sea surface emittance is mapped as a 

 series of gray shades on 9" photographic film (figs. 5-10), with a 1 .6°C temperature 

 resolution. The sensor records the amount of radiation being emitted by the sea surface 

 minus atmospheric attenuation caused chiefly water vapor. The water vapor of the entire 

 atmospheric column absorbs the sea surface radiation and reradiates at a lower temperature 

 (assuming bulk of water vapor is colder than sea surface). 



The sea surface temperatures detected by this sensor are from the top most radiating 

 layer of the sea surface of less than 0.1 mm in thickness (Clark, 1967). Sea surface effects 

 and the temperature gradient in the upper few millimeters of the sea may significantly 

 influence the detected temperatures in comparison with temperatures measured using con- 

 ventional surface measurement methods. Clark reports that vertical temperature drops of 

 from 0.5°C to 1 .0°C across a few millimeters of water are not uncommon and that behind 

 a weather front the drop may be as large as 10°C. However, infrared "skin temperatures" 

 versus those obtained from the 3 meter depth ships hull mounted sensor during an experiment 

 in Oct-Nov 1972, resulted in a mean temperature difference of only .31 °C in a sea state 

 2 (Beaufort scale) conditions. Slicks, sea state, foam, and spray also reduce infrared 

 emittance with respect to the representative sea surface temperature. However, the most 

 serious errors are introduced by clouds, smaller than the resolution of the sensors, which 

 contaminate the data but go undetected. Then, surface observations or higher resolution 

 remote sensing observations are necessary. Experiments comparing infrared effective 

 radiation temperatures with actual sea surface temperatures of the DAPP system under clear 

 sky conditions have shown differences ranging from 0-1 2°C, cooler than the surface 

 measurements. The errors were seen to be dependent on both the atmospheric conditions 

 and path length of the observation through the atmosphere (angular displacement of the 

 resolution element from the satellite subpoint). It is useful to observe that the spatial and 

 thermal resolution of the DAPP system in effect provides a "low pass filter" on sea surface 

 emittance. The scan spot values are effectively running average values of a small area 

 2-14 n.mi. in diameter, digitized into 1 .6°C temperature ranges. This is a very coarse 

 thermal resolution with respect to sensor sensitivity and calibration. The data provide a 

 synoptic field of sea surface temperature values for direct incorporation into oceanographic 

 operations, within minutes of the satellite overpass. The complicating factors affecting 

 absolute values and geographic gridding must be rigorously considered when machine pro- 

 cessing digital values at higher resolutions. Studies of the digital infrared data subjected 

 to computer enhancement and gridding will be published at a later date. 



